mcpField
S
skyvern

skyvern

github.com/davepoonAgent workflows3.7k73 installsUpdated today
Repository

About skyvern

Skyvern is a browser automation skill that uses natural-language actions and vision-based page interaction. It supports data extraction, secure login, and reusable workflows, and is intended for people automating browser tasks.

Key features

  1. 01Natural-language browser actions
  2. 02Schema-validated data extraction
  3. 03Credential vault with 2FA/TOTP
  4. 04Reusable parameterized workflows
  5. 05Session and multi-tab controls

Use cases

  1. 01Extracting product prices from a website
  2. 02Logging in and downloading an invoice
  3. 03Filling and submitting web forms
  4. 04Scheduling repeatable browser workflows

MCP support

Skyvern provides an MCP server with tools for browser sessions, multiple tabs, iframes, network inspection, and HAR recording. It can be configured as a Streamable HTTP server.

Install

Copy the command and run it in a terminal.

$npx skills add https://github.com/davepoon/buildwithclaude --skill skyvern

After installing, type/skyvernin Claude Code to confirm it loaded.

Details

From the project README

Skyvern Browser Automation

Control a real browser with natural language. Skyvern uses Vision LLMs and computer vision instead of brittle XPath/DOM selectors, so automations survive UI changes.

Setup

Cloud (recommended):

pip install skyvern
skyvern setup  # interactive client selection

Or add the MCP server directly to your Claude Code config:

{
  "mcpServers": {
    "skyvern": {
      "type": "streamable-http",
      "url": "https://api.skyvern.com/mcp/",
      "headers": {
        "x-api-key": "YOUR_SKYVERN_API_KEY"
      }
    }
  }
}

Get your API key at app.skyvern.com.

Key Features

  • Natural language actions — click, type, scroll, hover, drag-and-drop via plain English
  • Structured extraction — extract data with JSON schema validation and screenshot reasoning
  • Secure login — credential vault with 2FA/TOTP support (QR, email, SMS)
  • Reusable workflows — 23 block types, parameterized runs, cached scripts (10-100x faster on repeat)
  • 75+ MCP tools — session management, multi-tab, iframes, network inspection, HAR recording
  • Works with — Claude Desktop, Claude Code, Cursor, Windsurf, Codex

Usage

"Navigate to example.com and extract all product prices"
"Log into my account and download the latest invoice"
"Fill out the shipping form and click Submit"
"Take a screenshot of the current page"
"Build a workflow that runs this every Monday"

Links

Comments

Markdown supported